Semantics and Rhetoric

I’ve given this lecture the title of semantics and rhetoric, but for those of you who are already familiar with these terms and with the philosophical issues that are associated with them I have to confess that the title promises too much: I won’t actually be dealing with how a semantic theory could account in one fell swoop for all the facts of rhetoric. Here the term ‘semantics’ simply refers to the meanings of words, and I use it here as a way of organizing the discussion of some phenomena that are important in the interpretation of arguments, and in the evaluation of them. The term ‘rhetoric’ is my catchall term for the miscellaneous techniques of argument presentation that are not otherwise covered – and which look like the sort of thing that a rhetorician might be interested in. So, it’s not so much semantics-and-rhetoric together, as, first, some semantical stuff, and then some rhetorical stuff.

Semantics

So let’s talk about some of the ways that an awareness of the meaningfulness of words can be significant in the interpretation of arguments. Now, you might think that the meaning of words and sentences is what we’ve been talking about for, lo, these many weeks now, and in a sense that’s right. But for the most part we’ve accepted that we understand what the words and phrases in an argument mean without too much quibbling, and what we’ve really been concentrating on is what role those words and phrases – with the assumed meanings – have been playing in the arguments. In what follows we’ll be wondering things like ‘what does that word mean in that place?’ because to identify and/or evaluate the argument under consideration we do often need to clarify crucial terms.

Vague Terms - Indeterminate Meaning

Some terms in an argument are such that correct limits to their application are indeterminate or fuzzy.

        E.g. What exactly counts as 'bald'? How little money do you have if you are poor?

This can sometimes lead to problems.

      E.g. 'John is poor so he eligible for income support'

Given that you must be quite impoverished to qualify for income support, we really need to know how poor John is before we can evaluate the argument. 'Poverty' is too vague to enable evaluation.

You may recognise in this the origin of the Sorites[1] fallacy that I didn’t get time to look at last time. Let’s look at it now.

A sorites fallacy is one in which it is claimed that two alternatives cannot be uncontroversially distinguished, and that therefore there is no difference between them.

The argument appears to rely upon a hidden premiss that if two concepts are different that there is a sharp division between them.

                Property P₁ is defined in terms of X.

                Property P₂ is defined in terms of X.

                X can vary by small increments.

A has property P₁.

[For anything, B, with property P₁, if property P₁ is different from property P₂ then there is a point at which incremental variation of X from that which defines P₁ towards that which defines P₂ is sufficient to result in B having property P₂ rather than P₁.]

If A has P₁ and X is incrementally varied then A still has property P₁.

Therefore, there is no point at which incremental variation of X from that which defines P₁ towards that which defines P₂ is sufficient to reÿÿlt in B having property P₂ rather than P₁.

Therefore, P₁ and P₂ are not different.

(That’s looks pretty hairy itself. It makes more sense if you substitute:

  P₁ = ‘hairy’, P₂ = ‘bald’, X = ‘number of hairs’, A = ‘Bob’) 

Arguments like this are fallacious if the hidden premiss is not true (because the concept is genuinely vague), or if it is not obviously true and yet is not argued for.

Examples:

i.                     From the moment of conception, a foetus grows and changes constantly. At each moment it differs in only the slightest way from the previous moment. What, then, is the exact point at which the foetus ceases to be a mere foetus and becomes instead a person? For any point identified as the moment when a foetus becomes a person, the foetus at that point will not differ significantly from the foetus a moment earlier. Since we can’t identify any moment at which a foetus becomes a person, and yet the child is already a person when born, the foetus must be a person from the moment of conception.

ii.                   All animals have rights and we ought to respect these rights. How so? Humans have rights, we all accept that. But what is the relevant difference between humans and higher apes? – say the chimpanzee? Humans and higher apes are both conscious, intelligent beings, capable of learning and communicating through language. Apes, in turn, are closely related to other higher mammals, higher mammals to lower mammals, and lower mammals to other animals. Exactly where do rights come into the picture? We cannot draw a sharp and non-arbitrary distinction between those animals whose rights we ought to acknowledge and those we needn’t bother with, so the rights of all animals ought to be respcted.

iii.                  (Robert Nozick (1974), Anarchy, State, and Utopia, New York: Basic Books, pp 290-292)

1.                There is a slave completely at the mercy of his brutal master’s whims. He is often cruelly beaten, called out in the middle of the night, and so on.

2.                    The master is kindlier and beats the slave only for stated infractions of his rules (not fulfilling the work quota, and so on). He gives the slave some free time.

3.                    The master has a group of slaves, and he decides how things are to be allocated among them on nice grounds, taking into account their needs, merit, and so on.

4.                    The master allows his slaves four days on their own and requires them to work only three days a week on his land. The rest of the time is their own.

5.                    The master allowes his slaves to go off and work in the city (or anywhere they wish) for wages. He requires only that they send back to him three-sevenths of their wages. He also retains the power to recall them to the plantation if some emergency threatens his land; and to raise or lower the three-sevenths amount required to be turned over to him. He further retains the right to restrict the slaves from participating in certain dangerous activities that threaten his financial return, for example, mountain climbing, cigarette smoking.

6.                    The master allows all of his 10,000 slaves, except you, to vote, and the joint decision is made by all of them. There is open discussion, and so forth, among them, and they have the power to determine to what uses to put whatever percentage of your (and their) earnings they decide to take; what activities legitimately may be forbidden to you, and so on.

7.                    Though still not having the vote, you are at liberty (and are given the right) to enter into the discussions of the 10,000, to try to persuade them to adopt various policies and to treat you and themselves in a certain way. They then go off to vote to decide upon policies covering the vast range of their powers.

8.                    In appreciation of your useful contributions to discussion, the 10,000 allow you to vote if they are deadlocked; they commit themselves to this procedure. After the discussion you mark your vote on a slip of paper, and they go off and vote. In the eventuality that they divide evenly on some issue, 5,000 for and 5,000 against, they look at your ballot and count it in. This has never yet happened; they have never yet had occasion to open your ballot. (A single master also might commit himself to letting his slave decide any issue concerning him about which he, the master, was absolutely indifferent.)

9.                    They throw your vote in with theirs. If they are exactly tied your vote carries the issue. Otherwise it makes no difference to the electoral outcome.

The question is: which transition from case 1 to case 9 made it no longer the tale of a slave?

Ambiguity - Multiple Meanings

Ambiguity is, roughly, the phenomenon of multiple meaning. When something is uttered or written it is often said to be ambiguous when it admits of multiple interpretations or admits of multiple meanings. It can arise with regard to utterances or the words of phrases uttered.

More exactly, there are two senses in which we might describe an utterance or word or phrase as ambiguous:

1.                An utterance, word or phrase is ambiguous(1) if and only if it has more than one meaning.

                   E.g.                'bank'; 'found by some cows'; etc.

2.             An utterance, word or phrase is ambiguous(2) in a given context C if and only if it is misleading or potentially misleading because it is difficult to tell which of a number of possible meanings is intended in context C.

                   E.g.                the term 'person' in a debate on abortion; 'family' in public funding debates 

·     Notice that an utterance, word or phrase is ambiguous2 only if it is ambiguous1 — for multiple meanings to constitute a problem (ambiguity2) there must be multiple meanings (ambiguity1).[2]

·     We shall generally be interested in ambiguity as a defect, so the relevant sense is sense-2. It is only when multiple meanings are potentially misleading that fallacies arise.

·     multiple meanings can cause problems in even determining what the argument under consideration is.

        — pronouns

                E.g.         'John is still sick so he will be charged'

                        Does 'he' refer to 'John' or someone else? E.g. is the argument:

                        John is still sick                  or                    John is still sick                 

                        John will be charged                                 John's doctor will be charged

                We should be careful to fill out pronoun expressions since, once out of the context of         the surrounding discussion, their meaning might be unclear.

                In fact, words like 'here', 'now', 'he', 'I' should all be replaced by what they refer to.

        — scope of claim

                E.g. 1         'The early deaths of Joplin and Hendrix show how really dangerous drugs are'

                        Is this argument:

                        Joplin and Hendrix died early (from drug-overdose)    or   Joplin and Hendrix ...

                        Some drugs are really dangerous                                     All drugs are ...

                E.g. 2        'Judges voted Volkswagen the best car of 2001'          
                       
                Did all judges vote in this way or just some?

Given this understanding of ambiguity we can now make a further distinction based on the observation that because the meaning of a phrase can be seen as arising from the meanings of the individual words therein and the way they are structured, both individual word-meaning and structure can be a possible cause of ambiguity. The obvious consequence of this is that ambiguity must come in two forms.

1.             Verbal or Semantic Ambiguity (Lexical Ambiguity)

Most of the examples of ambiguity that we usually identify are examples of words that are ambiguous. This is generally referred to as semantic or lexical ambiguity. Another example is the claim:

                'OJ Simpson had his golf clubs taken from him to fund compensation for his wife's family'

Does this mean he had his wedge, five-iron, and other "ball-hitting implements" taken, or did he have the profit-making organisations of which golf players are members taken?

                   Fallacy of Equivocation (again)

You’ll probably remember that we had a quick look at how this sort of ambiguity can affect arguments last week in the section on ‘fallacies’. There we said that an argument involves the fallacy of equivocation when it exploits the different meanings of a word in different parts of the argument (typically, in different premises) to make the argument appear good when it is not.

E.g.         A simple, clear example from the Text is:

                Six is an odd number of legs for a horse

                Odd numbers cannot be divided by two               

                Six cannot be divided by two.

The word 'odd' is used in two senses — odd1 = unusual; odd2 = not even — and the argument gains any strength it has by ignoring this. The argument might be clarified as follows:

Six is an odd1 number of legs for a horse             Six is an odd2 number of legs for a horse

Odd2 numbers cannot be divided by two                OR                Odd2 numbers cannot be divided by two 

Six cannot be divided by two.                                                   Six cannot be divided by two.

·     With the meanings of the terms clarified so as to make the premises true (see the argument above right), there is little risk of thinking that they provide any support for the conclusion since the argument obviously equivocates.

·     With the meanings of the terms clarified so as to make the conclusion follow from the premises (see the argument above right), there is little risk of thinking all the premises are true.

E.g.         Less obviously, consider the following argument, seemingly implicit in remarks by a recent Prime Minister, in defence of the particular way spending cuts were distributed in a recent Budget. Cuts were made equally across all

                (A)                I cannot be criticised for fairly distributing a necessary burden

                (B)                The budget cuts were a necessary burden fairly distributed            

                                      I cannot be criticised for the budget cuts

                The notion of a 'fair distribution' is ambiguous, meaning:

                                1.                distributed equally across all

                                2.                distributed across all according to their capacity.

                For the argument to be valid it must use the notion in one sense throughout, BUT:

                — if sense 1 is the intended sense then premise B looks true but premise A seems                                 vulnerable. (Fairness might, arguably, require each person contribute according to                                          their capacity; the rich should, on this view, contribute more than the poor.)

                — if sense 2 is the intended sense then premise A may be thought true (for the reasons just                          given) but (given the way cuts were actually effected) premise B now seems                                     vulnerable.

·     With the meanings of the terms clarified so as to make both premises true, there is little risk of thinking that they provide any support for the conclusion since the argument obviously equivocates.

·     With the meanings of the terms clarified so as to make the conclusion follow from the premises, there is little risk of thinking all the premises are true.

                   General Strategy for Diagnosis

Where one suspects equivocation within an argument:

 (a) spell out the various senses of the term involved

 (b) restate the argument so that no equivocation occurs

 (c) evaluate each clarified argument. (If there is a fallacy of equivocation, none will be compelling.)

2.                Grammatical or Syntactic Ambiguity (Amphiboly)

A phrase can be ambiguous due to the structure of the phrase. Such cases are referred to as cases of syntactic ambiguity or amphiboly.

E.g.         —                Wartime poster:                                    SAVE SOAP AND WASTE PAPER

                —                 Definition of anthropology:                THE SCIENCE OF MAN EMBRACING WOMAN

                —                Newspaper headlines:                       FLYING PLANES CAN BE DANGEROUS

MAN BLOWS OUT HIS BRAINS AFTER TAKING AFFECTIONATE FAREWELL OF HIS FAMILY WITH A SHOTGUN

                —                Philosophy:                                          EVERYTHING MUST HAVE SOME CAUSE

Fallacy of Amphiboly

An argument is said to commit the fallacy of amphiboly just in case its apparent strength rests on reading a syntactically ambiguous phrase in the premises in one sense to ensure their truth and another sense to support the conclusion. [Cf Sainsbury, Logical Forms, Ch 4]

E.g.         He was found in the ditch by some cows                Everything must have some cause

Cows can find people.                                    Some one thing must be the cause of everything

General Strategy for Diagnosis

Where one suspects amphiboly within an argument:

 (a) spell out the various senses of the syntactically ambiguous phrase involved

 (b) restate the argument so that no equivocation occurs

 (c) evaluate each clarified argument. (If there is a fallacy of amphiboly, none will be compelling.)

Jargon - Unknown Meaning

Arguments often involve jargon-terms — that is, terms with specialised meaning in a particular area.

E.g.         'materialism' as used by philosophers means something quite specialised and something different from another common use of the term to describe someone who likes material comforts.

E.g.                 The notion of 'clear and distinct understanding' as used by Descartes means something quite specialised (scholarship is required to discern meaning).

Consequently, it can be difficult to evaluate an argument if it contains such a term whose meaning you do not know. This problem is solved by finding (by scholarship) or giving (by stipulation) a clear meaning to the jargon term so that all parties to the argument can see what is meant.

Definitions

To avoid problems surrounding the use of ambiguous terms in reasoning or terms whose meaning is simply unclear, one can — and, in disciplines like philosophy, often does — define a clear and unambiguous meaning for a term or phrase. Definitions are of various kinds.

(The book gives a list, but here is one which, I think, is clearer.)

In the first instance we can broadly divide kinds of definitions into those which are used to describe current meanings (lexical definitions) and those which are used to attach new meanings (stipulative definitions).

A.            Lexical Definitions

                Definitions which report current, conventional meanings for terms

Examples:

1.                Dictionary Definitions

Such definitions generally provide statements of conventionally accepted meanings of words which, though suitable for everyday purposes, are somewhat loose, inexact or incomplete. (Accordingly, the problem with their use in academic writing, to support a particular interpretation of a term in argument (say), is that the subtle shades of meaning on whichyourargument might depend may either be unspecified by the dictionary, or specified according to conventional usage but conventional usage might be misguided on the matter). [See Cederblom, p. 199 for inadequacies of such definitions.]

                Other kinds of definition which can also be used to give lexical definitions include:

2.                Denotative or Ostensive Definitions

                Giving the collection or class of things to which the term may be correctly applied, or                 paradigmatic examples of things to which it applies.

                E.g.                 By Eucalypts I mean Red Gums, Coolibahs, etc. (listing eucalypts)

                                By mental events I mean things like desires and beliefs.

3.             Logical Definitions

                Giving the set of conditions met by all and only those objects to which the term applies.

                I.e. giving the set of criteria necessary and sufficient for the application of the term.

                E.g.                By a square I mean any equal sided rectangle.

                                Something is a square if and only if it is an equal sided rectangle.

                A good logical definition is neither too broad nor too narrow; is not circular; uses clearly                 understood terms; is not negative when it can be affirmative.

The foregoing kinds of definition might be employed to report the current meaning of a term. On the other hand, definitions might be used to give new meaning to a term.

B.                Stipulative Definitions

                Attaching unconventional, new meaning to a term.

                Ordinarily one stipulates a meaning when:

                (a) one believes that a word is ambiguous and seeks to forestall any possible equivocation; or

                (b) one finds that no word exists for a concept to be used.

                E.g.                (a) "For the purposes of this debate let us use democracy to mean ..."

(b) "Let bleen be that property that something has just if it is blue until the year 2000 and green thereafter."

Examples:

                Denotative or Ostensive Definitions

                (see above)

                Logical Definitions

                (see above)

                A potentially problematic aspect of stipulative definition is:

                Persuasive Definition

                       Attaching a different literal meaning to a word while preserving its old emotional or                 evaluative impact.

E.g.         Defining the "true" Australian as someone who cares about the plight of farmers; or defining "right-thinking people" as those against fluoridation of drinking water.

Sometimes, rather than eliminating ambiguity, this type of definition reintroduces it, leading to the problem of redefinition. That is where someone uses the persuasive definition of a term so that it applies to a desired group yet has all the emotional or evaluative impact of the term as originally understood.

                E.g.                 All right-thinking people are against fluoridation

                                  Reasonable people are right-thinking                                       

                                Reasonable people are against fluoridation

The first premise of this argument is true given the above persuasive definition. (It is in fact a truism, amounting to the claim "All people against fluoridation are against fluoridation"). However, in this sense of "right-thinking person" the second premise is questionable. This premise might appear true because we may be tempted to attach the original evaluative impact to the term "right-thinking person".

The argument appears compelling only if one equivocates between the new stipulated meaning of "right-thinking person" and the original meaning of the phrase.

Rhetoric

So much for that short treatment of how the ‘meanings’ of words and phrases can be abused in arguments. Now we shall turn to those techniques that don’t fit quite so neatly into such a classification but which are nevertheless of some significance in affecting the effectiveness of arguments. I call these things ‘rhetoricisms’ for my own convenience, but you’re quite welcome to reject this term as an impossibly ugly neologism.

Irrelevancies

Text or conversations containing arguments almost always contain a great deal which is not part of the contained argument. Some of it might be by way of introduction, helpful asides, discounting, etc. Much of this is useful in the presentation of an argument. However, when it comes to analysing an argument, this material is often dropped since it does not constitute part of the argument's premises, intermediary conclusions, or overall conclusion.

There are probably an infinite number of ways in which we can imagine that parts of a text being presented to us as an argument are irrelevant. Here are just a couple which seem to be particularly popular:

a.                    going off on tangents to gain argumentative advantage.

This often appears upon closer examination to be merely excessive verbiage.

E.g.                Q: 'Should the government's tax changes be accepted?'

A: 'The government has a lousy record on tax change.

These changes are unnecessary because […]'

The argument is simply:

                                […]                                                                                                        

                                The government's tax changes are unnecessary

An extreme case is where someone is asked why we should believe their claim X, and in fact no argument is offered at all! The point is avoided, not addressed. When the excess verbiage is eliminated as a mere tangent, nothing remains.

b.                    repetition

E.g. '… because it is wrong; it's immoral; it's obscene!'

In extreme cases repetition actually takes the place of argument. When repetitious claims are eliminated, sometimes only a single assertion of the very point at issue remains. This is a commonly employed rhetorical device (i.e. a device for producing good rhetorical argument). The audience may as a matter of psychological fact become convinced of some claim A if its assertion is repeated often enough, even in the absence of good reasons or argument for believing it.

Assuring, guarding and discounting are three strategies commonly used to secure acceptance of premises or reasons as true.

Assuring

·     Presenting reasons that are "assuredly true" ... to convince a disbelieving audience.

                — by citing authority as an assurance                    E.g.                "Economists agree that ..."

                — by showing you yourself accept them                       E.g.                "There is no doubt that ..."

                     (conversationally implying there are reasons)

                — by citing obviousness                                                         E.g.                "Clearly ..."

                     (implying that disbelief amounts to ignorance)

In honest argumentation assurances may save time by referring audience to further argument not presented but available.

In dishonest argumentation they may be used to cover weak points in an argument.

Irrelevant assuring

E.g. 'You obviously cannot play golf in Alaska in January so there's no point in bringing your clubs'

                                        You cannot play golf in Alaska in January

                                       There's no point in bringing your clubs

But assuring is not always irrelevant.

E.g. 'You obviously cannot play golf in Alaska in January so there should be no thought of your bringing your clubs'

                                                You obviously cannot play golf in Alaska in January

                                               There should be no thought of you bringing your clubs

Guarding

·     Presenting reasons that are suitably weakened by ...

                —                restricting scope or extent of claims - less general  

                                E.g. "Some ..." instead of "All ..."

                —                retreating from certain claims to merely probable claims - less certain                           

                                E.g. "Evidence suggests that ..." instead of "It is plainly the case that ..."

                —                retreating from knowledge-claims to belief-claims - less dogmatic

                                E.g. "I now think that ..." instead of "I know that ..."

In honest argumentation guarding may yield reasons which are more believable.

Don't commit yourself to claims stronger than you need ... but don't go too weak — find a "middle way".

In dishonest argumentation guarding often functions to cover the reasoner by only making explicit a claim that can be defended whilst perhaps encouraging the audience the accept the conversationally implied stronger claim. (Anecdotal evidence is often used in this way. One explicitly says that people have suggested that X, conversationally implying X.)

Irrelevant guarding

        E.g. 'I think Miranda is at home so we can meet her there'

                                        Miranda is at home     

                                      We can meet her there

But guarding is not always irrelevant to the argument.

        E.g. 'I think Miranda is at home so I don't expect to see her at University today'

                                        I think Miranda is at home                                     

                                        I don't expect to see her at University today

Discounting

·     Dispelling doubts surrounding reasons by citing possible criticism only to discount it.

E.g. 'Since historically public debt leads to inflation, I maintain that, despite recent trends,                       inflation will return' [from Exercise VI #2, Ch 2]

          The argument is simply:

                                                Historically public debt leads to inflation 

                                                Inflation will return

E.g.   The ring is beautiful but expensive [… so let's not buy it.]          (Discounting its beauty)

           The ring is expensive but beautiful [… so let's buy it.]          (Discounting the expense)

A but B                                 A although B

          (i)          asserts A                                          (v)          asserts A

          (ii)          asserts B                                         (vi)          asserts B

          (iii)          suggests A/B opposition            (vii)          suggests A/B opposition

          (iv)          A discounted by B                       (viii)          B discounted by A

                    Discounting A                      Discounting B

                    A but B                                                A although B

                    A however B                                       A (even) though B

                    A nonetheless B                                 A even if B                   

·     Discounting is useful for:

                (a)                pre-empting moves by opponent

                (b)                blocking conversational implication

                (c)                avoiding side issues or tangents

·     Discounting can be abused by:

                (a)                pre-empting easy objections to hide or bury difficult ones

                (b)                using it to imply that opponent holds some crazy view (by means of (iii)                                      above - the suggested opposition)

              E.g.                I agree that pollution is bad but stopping all industry won't work.

Evaluative Language (vs Descriptive)

The use of evaluative language is often important in pushing an argument through and for that reason we must be aware of its use.

·     Evaluative terms like 'good', 'proper', 'efficient' or 'beautiful' are typically said to invoke (positive, virtuous) standards against which "things" (e.g. objects, events, etc.) can be evaluated. Since such evaluations go beyond a mere description of how things are, they can be contested in a way that mere description cannot be. You can disagree with someone who describes a particular act as one that is good in a way that you cannot if they merely report the act itself.

        The standards of evaluation can be vague, and may vary culturally but imprecision and variation do not show that there are no shared standards, that it is merely a matter of taste.

·     Recognising evaluative use of language is important but can be difficult.

        Test: Does the use mean something is good or bad; right or wrong; ought be done or ought not.

        E.g. 'homicide' vs 'murder'

                'refugee' vs 'illegal immigrant'

                'removed' vs 'stolen'

Euphemism and Spin Doctoring

·     Euphemism — The use of language to describe something in more gentle or favourable terms.

        E.g. "It's only downsizing, so it's not anything to be concerned about."

·     The euphemism should in many cases be scrutinised.

Figurative Language

·     Language is not always used in its literal sense but metaphorically.

        E.g. a glaring mistake

        Arguments may sometimes contain figurative language where the conclusion is backed up by reliance on some metaphor.

        E.g.

                                        The High Court is a court of autocratic kings

                                        We can and should ignore their decisions

·     Criticism naturally focuses on the appropriateness of the metaphor